The present invention relates generally to the field of an apparatus for calculating an inverse of an element in a finite field.
A finite field is a field that contains a finite number of elements, which is widely used in various engineering fields. The finite field inverse operations can be roughly divided into four types: an inversion based on Fermat's Theorem, an inversion based on Extended Euclid's Algorithm, an inversion based on Montgomery Algorithm and an inversion based on look-up table technique.
All kinds of finite field operations can be effectively applied in various cryptographic applications and coding techniques. The designing of an effective finite field inverse operation plays a key role in cryptosystem implementation. Many well-known finite field inverters in the prior art, including software inverters and hardware inverters, have some shortcomings, for example, the performance cannot reach the requirement of high-speed, small-area and low-power consumption.
In order to overcome the deficiencies of the prior art, the object of the present invention is to provide a finite field inverter, which uses a search tree structure to calculate an inverse of an element in a finite field; the present invention is more efficient when compared with the existing finite field inverter, and it has the characteristics of high-speed, small-area and low-power consumption in inverse operation over finite field GF(2n).
The object of the present invention is achieved by the following technical solutions:
A finite field inverter includes:
an input port, configured to input an operand a(x);
a search tree inverse circuit, configured to perform an inverse operation of an operand a(x) in the finite field GF(2n) based on a search tree structure; and
an output port, configured to output an inversion result b(x) of the operand a(x).
The search tree inverse circuit is provided with a left search tree and a right search tree. The left search tree and the right search tree both include tree nodes for processing inverse operations over the finite field GF(2n) and connecting wires between the tree nodes. The tree nodes include a root node, internal nodes and leaf nodes. Each path from the root node to a leaf node represents an element in the finite field GF(2n). The connecting wires between the tree nodes connect the path representing the operand a(x) with the path representing the inversion result b(x).
In the search tree inverse circuit, the operand a(x) is represented by a path from the root node to a leaf node n1, the inversion result b(x) is represented by a path from the root node to a leaf node n2, and the connecting wire is provided between the leaf node n1 and the leaf node n2.
In the search tree inverse circuit, the tree nodes include a NXOR logic gate, a AND logic gate and a data selector MUX. One input of the NXOR logic gate is the bit value of the operand a(x), and the other input is i2. One input of the AND logic gate is i0, and the other input is the output of the NXOR logic gate. The data selector MUX is provided with a data input i2, a gated input i3 from its child node, a inverse output o2 and a output o3 transmitted to its parent node. If the tree node is a root node, i0 is 1; if the tree node is an internal node or a leaf node, and at the same time also a left child node, i2=0; if the tree node is an internal node or a leaf node, and at the same time also a right child node, i2=1; if the tree node is a root node or an internal node, the AND logic gate outputs the result of Boolean AND operator to its child node; if the tree node is a leaf node, the AND logic gate outputs the result of Boolean AND operator to a leaf node connected thereto; for the root node of the left search tree, i2=0, and for the root node of the right search tree, i2=1.
The operand a(x) and the inversion result b(x) have the following forms:
a(x)=an-1xn-1+an-2xn-2+ . . . +a0;
b(x)=bn-1xn-1+bn-2xn-2+ . . . +b0.
Compared with the prior art, the present invention has the following advantages and technical effects:
The present invention achieves finite field inverse operations by a structure of search tree; compared with the existing finite field inverter, the present invention has advantages of high-speed, small-area and low-power consumption in a finite field GF(2n) inverse operation, and therefore, it can be applied to various engineering fields, particularly to the hardware implementation of cryptographic algorithms and to the solving of various mathematical problems.
Further characteristics and advantages of the present invention will be apparent upon reading the following description provided by way of non-limiting examples, with reference to the attached drawings.
As shown in
The following paragraphs will describe in detail the components of the inverter according to the present invention.
1. Input port: as shown in
a(x) can be expressed as the following form:
a(x)=an-1xn-1+an-2xn-2+ . . . +a0
wherein an-1, an-2, . . . , a0 are elements in GF(2).
2. Output port: as shown in
3. Search tree inverse circuit: as a main component of the inverter, the search tree inverse circuit is a core component of the present invention, and it includes a plurality of tree nodes and connecting wires connected between the tree nodes.
As shown in
As shown in
i1 and i2 are two inputs of NXOR, i1=ai, i2=0/1; one input i0 of AND is from an output of its parent node, and the other input of AND is from an output of NXOR; o1=o2, o1 and o2 are outputs of AND, they are directly outputted to the left and right child nodes of the node. If the tree node is a left child node, the two inputs of NXOR are ai and 0; otherwise, the two inputs of NXOR are ai and 1. As shown in
The differences between the three kinds of tree nodes are input and output ports. For example, in the root node, i0=1, i1=an-1, i2=0/1, wherein an-1 is the value of the (n−1)th bit of the finite field element a(x), the NXOR logic gate carries out logic operation on the value an-1 of the (n−1)th bit of a(x) and the other input i2, and then outputs the arithmetic result to one input port of the AND logic gate, the AND logic gate carries out logical AND operation on the arithmetic result and 1, and then outputs the logical AND operation result respectively to the left child node and the right child node of the root node; in the internal nodes, i1=ai, i2=0/1, wherein ai is the value of the ith bit of the finite field element a(x), the NXOR logic gate carries out logic operation on the value ai of the ith bit of a(x) and the other input i2, and then outputs the arithmetic result to one input port of the AND logic gate, the AND logic gate carries out logical AND operation on the arithmetic result and the output i0 from the parent node of the internal node, and then outputs the logical AND operation result respectively to the left child node and the right child node of the internal node; in the leaf node, i1=a0, i2=0/1, wherein a0 is the value of the 0th bit of the finite field element a(x), the NXOR logic gate carries out logic operation on the value a0 of the 0th bit of a(x) and the other input i2, and then outputs the arithmetic result to one input port of the AND logic gate, the AND logic gate carries out logical AND operation on the arithmetic result and the output i0 from the parent node of the leaf node, and then outputs the logical AND operation result to the leaf nodes connected to the said leaf node. If the tree node is a left child node, i2=0; and if the tree node is a right child node, i2=1. The present invention includes two search trees, that is, a left search tree and a right search tree, for the root node of the left search tree, i2=0, while for the root node of the right search tree, i2=1.
The data selector MUX is defined as follows: i2 is a data input and i3 is a strobe input which comes from the input of the child node of the node; o2 and o3 are outputs, o2 is output to the output port of the inverter that is a part of b(x), and o3 is output to the parent node of the node. If o3=i3, and only if i3=1, o2=i2, that is to say, if the output from the child node of the node is 1, the output of the inverter at the node is the input i2 of the NXOR logic gate.
Here is an example with n=4, by which the work process of the inverter according to the present invention is illustrated.
Firstly, as shown in
Suppose the to-be-inversed element a(x)=x, which can be expressed as (0010)2 in binary. Because an-1=n1=1, the n1 node is in the correct path, the AND logic gate outputs 1 to the left and right child nodes. In the n2 node, because i1=0, i2=0, i0=1, the AND logic gate outputs 1 to the left and right child nodes, that means, the n2 node is also in the correct path. It can be obtained by the same way that the n3 node and the n4 node are all in the correct path, so the path from n1 to n4 represents (0010)2, namely, a(x). Because the n4 node is connected with the n5 node, the output of the AND logic gate of the n4 node is transmitted to the n5 node. That is to say, the inverse of a(x) is in the finite field.
Secondly, as shown in
Because the n4 node is connected with the n5 node, the output value 1 of the n4 node is transmitted to n5 node. The n5 node can directly transmit the value 1 to its parent node n6 and its ancestor nodes n7 and n8. Meanwhile, n5 node, n6 node, n7 node and n8 node output the internal data (namely the input i2 of the NXOR logic gate) to the output port, namely said internal data is b(x). Therefore, b(x) is the inverse of a(x).
The above examples are preferred embodiments of the present invention, but the present invention is not limited to said embodiments; any modification, replacement, combination and simplification may be made to the embodiments without departing from the scope of the invention.
Number | Date | Country | Kind |
---|---|---|---|
2012 0 275733 | Aug 2012 | CN | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/CN2012/085948 | 12/5/2012 | WO | 00 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2014/026451 | 2/20/2014 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
4975867 | Weng | Dec 1990 | A |
6138133 | Oh | Oct 2000 | A |
6779011 | Weng et al. | Aug 2004 | B2 |
6862354 | McGrew | Mar 2005 | B1 |
20030198343 | Morioka | Oct 2003 | A1 |
20030219118 | Beverly | Nov 2003 | A1 |
20040107341 | Hall | Jun 2004 | A1 |
20100306299 | Reidenbach | Dec 2010 | A1 |
20130173917 | Clifton | Jul 2013 | A1 |
Number | Date | Country |
---|---|---|
1032595 | Apr 1989 | CN |
101572602 | Nov 2009 | CN |
Entry |
---|
Extended European Search Report, Application No. EP12879153.0. Date of mailing: Mar. 3, 2015. European Patent Office, Munich, Germany. |
International Search Report and Written Opinion, English translation of, International application No. PCT/CN2012/085948. Date of mailing: May 16, 2013. SIPO, Beijing, China. |
Guo, J-H. et al., “Systolic Array Implementation of Euclid's Algorithm for Inversion and Division in GF(2m).” IEEE Transactions on Computers, vol. 47, No. 10, pp. 1161-1167, Oct. 1988. |
Sarkar, P., et al., “A Parallel Algorithm for Computing Simultaneous Inversions with Application to Elliptic Curve Scalar Multiplication.” IEEE Midwest Symposium on Circuits and Systems, vol. 2, pp. 782-785, Dec. 2003. |
Number | Date | Country | |
---|---|---|---|
20150067011 A1 | Mar 2015 | US |